JEC COMPOSITES MAGAZINE - Issue #81 - May/June 2013 - 61

Existing cost models
Known cost models are often part-specific, e.g. for composite
beams or wing-like structures [10], but there are some generic
software tools available. Most of them can estimate costs along
input parameters such as part complexity and production steps
and some allow process comparisons, but hardly any include
post-production operations such as non-destructive testing.
Three major families of cost prediction approaches can be
distinguished, each one with clear benefits, but also known
limitations.
Standard-based cost estimations
Models in this group can be used to estimate recurring costs by
estimating the required amount of labour time along a predetermined set of industrial engineering standards. Combined
with an hourly wage rate and material costs, this leads to fairly
accurate numbers. However, the integration of new manufacturing processes is not supported and non-recurring costs get
neglected [5], [6].
First-order models
Based on the theoretical work of the Massachusetts Institute of
Technology, such models use a sum of first-order equations for
each time step. While being limited on simpler parts, the support for new processes is good [3], [7], [8].
Knowledge-based models
Most industry-used models are based on existing knowledge
from comparable parts already in production. This allows very
accurate cost predictions, but is limited to existing manufacturing processes [9]. The main drawback of existing cost models is
their high input requirement, mostly for an accurate description
of the part complexity, which is most certainly not available at
early design stages.

Theoretical approach
The most likely boundary conditions for an early design stage
are a known envelope for the part, major force loadings and
connection points. A preliminary design is usually not optimized regarding part costs, but the potential effect on cost optimization within this stage is significantly greater than anywhere
else during part development and production.
A design-for-process approach tries to minimize the large cost
drivers for a manufacturing process, e.g. raw material costs or
tooling requirements.
For almost all parts within an aircraft, more than one manufacturing process is theoretically possible. Within the early design
stage, a part can often be optimized to perfectly match a certain
processing strategy. Therefore, an ideal pre-part cost estimator
does not necessarily need to produce exact figures in terms of
final part costing, but needs to highlight the best process for the
current design, and to show major limitations on the follow-up
processes, which can potentially be reduced or eliminated by
acceptable design changes. An alpha version of a new model for

early design stages has been developed for R&D programmes
carried out by FACC AG (FACE_ROAD: FACC Advanced
Cost Estimator for Research Or/And Development projects).
This cost calculation sheet for R&D projects works with a
limited set of input parameters that allow a rough preselection
along best-practice standards within the aerospace industry
(e.g. a 60m² part is not suitable for Resin Transfer Moulding
due to press limitations) and highlights the limiting factors
for the follow-up processes. After a literature study and expert
interviews, a set of important influence factors when selecting
processes for CFRP parts has been set up and integrated into a
cost calculator spreadsheet. With input numbers from the user,
the file automatically compares those inputs against technical,
economic or company-related limits. An excerpt of the input
parameters can be found in Table 1.
Tab.1: Influence parameters (excerpt)

Process influence

Company influence

Customer influence

Part count per annum

Experience

Material qualification

Projected area

Autoclave availability

Preferences

Surface tolerances

Press availability

Cost expectations

Minimum radii

Tape laying availability Qualification know-how

Curvature

Preferences

Part tolerances

Cost targets

% of UD fibres
Tooling complexity

Not all input parameters are hard-based facts. Limited knowhow or non-existing hardware may eliminate a certain process,
even if it might be an optimal solution for other manufacturers.
In order to achieve comparability between this range of various
influences, the software model is based on cost ratios. Cost
ratios are used fairly commonly within the aircraft industry to
allow a comparison of typically not directly related characteristics of a part. The most widely known example would be costto-weight ratios. Cost-to-weight ratios (Ctw ratios) are used
mostly to determine whether a lighter part is worth additional
costs or not. Ctw ratios can reach up to €1000/kg and are a key
factor for the success of CFRP reinforced parts in the aircraft
industry, but are heavily dependent on customer requirements
and certain part features.

Experimental work
FACE_ROAD has been used in an on-going research programme to test a process selection that was setup by experienced process engineers. The part is an annulus filler used to
close the gap between two fan blades on large passenger jet
engines. Due to the part rotating with the fan blades, it is highly
centrifugally loaded, while having to keep its shape to allow
proper engine functioning (Figure 3). A list of requirements
and constraints (excerpt) can be found in Table 2.
With said list of requirements, FACE_ROAD proposes RTM
No81 May 2013 /

jec composites magazine 61

JEC COMPOSITES MAGAZINE - Issue #81 - May/June 2013

Table of Contents for the Digital Edition of JEC COMPOSITES MAGAZINE - Issue #81 - May/June 2013